07.0 PP 103 145 What Is Persecution

3 What is “Persecution”?
T
he preceding chapters have articulated a political approach to
asylum, one that explains why asylum should be reserved for
persecuted people. Asylum provides not only protection to the
refugee, but also chastises the persecutory state for its misconduct. But to
precisely which kinds of harm does the label “persecution” apply? The
question is a difficult one. Certain kinds of harm obviously qualify: for
example, the murder or torture of political dissidents or ethnic or religious
minorities constitute the classic cases of persecution. But beyond this core,
which harms does the concept of “persecution” encompass? The UNHCR
Handbook, meant to provide interpretive guidance to courts and admin-
istrative agencies, begins its discussion of persecution with a striking
disclaimer: “There is no universally accepted definition of ‘persecution,’
and various attempts to formulate such a definition have met with little
success.”1 Guy Goodwin-Gill, a prominent commentator, concurs that
the definition of persecution has “no coherent or consistent jurispru-
dence.”2 American law, in particular, is in disarray. The asylum statute
leaves “persecution” undefined, and the agency charged with issuing
regulations – now the Department of Homeland Security – has offered
little regulatory guidance. The federal courts of appeals, meanwhile, have
employed a number of definitions which appear similar on their face but,
in fact, vary in subtle ways that can have important implications.3
The growing acceptance of a humanitarian conception of asylum partly
explains the confusion over persecution’s meaning. The term “persecu-
tion” seems to connote more than just the bare infliction of harm; perse-
cution is a particular kind of harm. Yet from a humanitarian standpoint,
the purpose of asylum is to protect refugees regardless of the particular
kind of harm to which they are exposed. To give “persecution” its plain
meaning would, from a humanitarian perspective, amount to a deviation
from asylum’s animating purpose. As the humanitarian approach has
taken root, this tension has been eased by broadening the definition of
“persecution” to encompass a wider array of types of harm. For example,
103
Cambridge Books
https://doi.org/10.1017/CBO9780511626845.004 Online
Published © Cambridge
online by CambridgeUniversity Press,
University Press 2010
104 Rethinking Asylum: History, Purpose, and Limits
in the United States, “government sanction” was initially regarded as a

necessary element of persecution; so too was a punitive intent.4 By the late
1990s, those prerequisites had been eliminated. The American trend has
been in the direction taken more definitively by Canadian, British, and
Australian courts: persecution means nothing more than serious harm
against which the state is unable, or unwilling, to provide protection.5
That gloss on “persecution,” however, is problematic for several rea-
sons. First, it cannot account for the requirement in the Convention
definition that persecution be “for reasons of race, religion, nationality,
membership of a particular social group, or political opinion.” From a
humanitarian standpoint, that requirement – known as the “nexus
clause” – introduces yet another departure from the purpose of asylum
to protect. Not only is it arbitrary to focus on persecution to the exclusion
of other harms, but it is also arbitrary to focus on certain reasons for
persecution to the exclusion of other reasons. While courts have found it
possible to gloss “persecution” in a way that promotes humanitarian
ends – defining it as the absence of state protection against serious harm –
they have had much more difficulty evading the nexus clause.
The humanitarian gloss on “persecution” is problematic in a more
fundamental way as well. It fails to recognize that persecution is a dis-
tinctive kind of harm that warrants a distinctive kind of response: con-
demnation and (if the persecution is sufficiently widespread) external
interference. The gloss I offer is grounded in the political approach to
asylum advanced in the preceding chapter. The broad theme is that “per-
secution” describes state conduct warranting international condemnation
and suspending a state’s sovereign right against external interference.
The persecution requirement, then, rests upon a theory of international
legitimacy – a theory that explains when such interference is justified.
I do not offer a comprehensive theory of international legitimacy.
Instead, my goal is merely to show how, with the concept of interna-
tional legitimacy as an anchor point, the persecution requirement
(together with the nexus clause) can be given a coherent and focused
meaning, and to illustrate that meaning by reference to a number of
difficult questions of application.
Persecution and international legitimacy

International law imposes a general rule against interference in the
internal affairs of other states. That rule stems from a commitment to
Cambridge Books
What is “persecution”? 105
the principle of communal self-determination, the idea that a political

community with a common way of life should have the freedom to
decide for itself how it should be governed. That principle follows from
the liberal idea that legitimate government must rest upon the consent of
the governed. And it is implemented by what I shall call a “presumption
of legitimacy,” the assumption that, as a general matter, governments
do indeed exercise power with the consent of their citizens. That pre-
sumption is reflected in the related doctrines of state sovereignty and
territorial integrity. Ordinarily, states have exclusive jurisdiction and
control over their territories, and external interference constitutes a
wrongful act of aggression.
Of course, except in unusual situations, no “social contract” is ever
actually ratified by citizens. Rather, as Michael Walzer explains:
Over a long period of time, shared experiences and cooperative activity of
many different kinds shape a common life. “Contract” is a metaphor for that
process of association and mutuality, the ongoing character of which the state
claims to protect against external encroachment. The protection extends not
only to the lives and liberties of individuals, but also to their shared life and
liberty, the independent community they have made, for which individuals are
sometimes sacrificed.6
Not every state acts with the tacit or express authorization of its
citizenry. Nonetheless, the presumption of legitimacy reflects a practical
judgment that communal self-determination is better served by a rule
prohibiting external interference rather than by a rule permitting it. Recall
Pufendorf’s comment that an international system in which other states
constantly “thrust [themselves] forward as a kind of arbitrator of human
affairs … could easily lead to great abuse.”7 Walzer explains:
[T]he recognition of sovereignty is the only way we have of establishing an
arena within which freedom can be fought for and (sometimes) won. It is this
arena and the activities that go on within it that we want to protect, and we
protect them, much as we protect individual integrity, by marking out bound-
aries that cannot be crossed, rights that cannot be violated. As with indivi-
duals, so with sovereign states: there are things that we cannot do to them,
even for their own ostensible good.8
The effect is to permit states to adopt a wide range of domestic institu-

tions and laws, free of external interference.
The presumption of legitimacy, however, is not absolute. There are
situations when “the violation of human rights within a set of
Cambridge Books
boundaries is so terrible that it makes talk of community or self-

determination … seem cynical and irrelevant, that is, in cases of ensla-
vement or massacre.”9 Recall Rawls’ concept of outlaw states and
Arendt’s concept of “rightlessness,” discussed in Chapter 2. When
such harm is widespread, a broad-gauged response by bystander states
is appropriate. As Walzer puts it:
If the dominant forces within a state are engaged in massive violations of
human rights, the appeal to self-determination … is not very attractive. That
appeal has to do with the freedom of the community taken as a whole; it has
no force when what is at stake is the bare survival or the minimal liberty of
(some substantial number of) its members … [W]hen a government turns
savagely upon its own people, we must doubt the very existence of a political
community to which the idea of self-determination might apply.10
In such cases, external interference advances rather than impinges on

the principle of communal self-determination.
When harm is serious but not so widespread, wholesale intervention
and overthrow would be inappropriate. That does not mean, however,
that bystander states are without recourse. Asylum is a means to remedy
on a retail level state misconduct that, if sufficiently widespread, would
justify more intrusive forms of external interference.
The elements of “persecution”

The meaning of “persecution” should be informed by the role of asylum as
an international sanction. “Persecution” is conduct so at odds with the
principle of communal self-determination that a presumption of legiti-
macy cannot be justified, and external interference is warranted. Such
mistreatment cannot be trivial; it must involve the infliction of serious
harm. The mistreatment must also be motivated by “illegitimate reasons” –
reasons inconsistent with the theory behind a presumption of legitimacy,
namely, that a state acts as the fiduciary of its citizenry. The nexus clause –
the requirement that persecution be on account of “race, religion,
nationality, membership of a particular social group, or political opi-
nion” – provides examples of such illegitimate reasons. A state that seeks
to inflict serious harm on a citizen because he or she is of a different race,
religion, nationality, or because of some other ascriptive characteristic, or
because he or she possesses a certain political opinion, cannot intelligibly
be said to act on behalf of that citizen. “Persecution” thus has three
Cambridge Books
elements: (1) serious harm that is (2) inflicted or tolerated by official agents
(3) for illegitimate reasons. In this chapter, I consider in greater detail the
first and third element; Chapter 4 will take up the second element.
Serious harm
One element of “persecution” is that it involves serious harm of some
sort. “Persecution,” as one court has put it, is an “extreme concept.”11 It
“does not encompass all treatment that our society regards as unfair,
unjust, or even unlawful or unconstitutional.”12 The presumption of
legitimacy affords states wide latitude to arrange their affairs as they
wish. Thus, asylum policy is not designed to remedy every injustice or
injury. Minor harms or inconveniences, even if carried out for illegitimate
reasons, do not call for international condemnation and interference.
It is easy to think of harms that clearly cross the threshold of requisite
seriousness: enslavement, torture, imprisonment, death. We can also
(perhaps with somewhat greater difficulty) identify a lower bound: per-
secution involves “more than a few isolated incidents of verbal harass-
ment or intimidation, unaccompanied by any physical punishment,
infliction of harm, or significant deprivation of liberty.”13 The threshold
of seriousness lies somewhere between “threats to life or freedom,” on the
one hand, and “mere harassment and annoyance,” on the other hand.14
Actions that “might cross the line from harassment to persecution
include: detention, arrest, interrogation, prosecution, imprisonment, ille-
gal searches, confiscation of property, surveillance, [or] beatings.”15
Often it is hard to decide whether treatment crosses the threshold of
seriousness, and judges can reasonably disagree. This difficulty is lar-
gely due to the impossibility of placing on a single spectrum the infinite
variety of incommensurable experiences of harm. The case of Li Xu
Ming is representative of this difficulty. Li was forcibly subjected by
Chinese population control officials to a half-hour long gynecological
exam to determine whether she was pregnant.16 When she tried to
resist, she was physically restrained by two nurses and threatened
with a sterilization operation.17
The majority held that the examination did not constitute persecu-
tion, contrasting Li’s case with more usual cases of arbitrary detention
and physical abuse. In Prasad,18 for example, a Fijian of Indian descent
was arrested by ethnic Fijians, placed in a cell for four to six hours, and
“hit on his stomach and kicked from behind.” The majority held that
Cambridge Books
Li’s forced examination was significantly less serious than Prasad’s

detention both in the intensity of harm inflicted and the duration of
that harm.19 The dissenting judge, meanwhile, compared Li’s forced
examination to rape, and rejected the length of detention as a reliable
measure by which to judge the seriousness of the harm inflicted.20
“Seriousness” is a judgment call, and different judges might reason-
ably reach very different answers. Precedent serves as a guide to judges
in deciding whether harm passes the requisite threshold of seriousness,
but cannot constrain them entirely: factual variations are endless –
suppose Prasad had only been hit on his stomach? or arrested twice
for four hours each, but had never been hit? and so on – and we lack the
means for easy comparison. Consequently, as one court has put it,
“courts have tended to consider the subject on an ad hoc basis.”21
Illegitimate reasons
Seriousness is not the only element of persecution. After all, states routi-
nely inflict harm that we might regard as persecutory if done for the
wrong reasons, but that we regard as entirely proper when done for the
right reasons. Most obviously, states imprison people. Imprisonment is
persecutory when used, say, to punish a religious minority for its beliefs;
but it is legitimate when used to punish a murderer. States also institu-
tionalize people and force them to undergo medical treatment. Shock
treatments might be regarded as legitimate when given to treat the
mentally ill, but illegitimate when given to “treat” homosexuals or others
regarded as “socially deviant.” States also take people’s property in the
form of taxation and condemnation; those legitimate takings are distin-
guished from illegitimate confiscation.
Thus, another element (and the more important one, for present
purposes) is that the serious harm be inflicted for illegitimate reasons –
reasons inconsistent with the usual presumption that a state acts as the
fiduciary of its citizenry. As Judge Richard Posner has written, persecu-
tion is “punishment or the infliction of harm for political, religious, or
other reasons that this country does not recognize as legitimate.”22 The
definition of “persecution” is, therefore, undergirded by a political
theory that describes which reasons for inflicting official harm are
legitimate and which are illegitimate.23
The nexus clause – which requires that persecution be “for reasons of
race, religion, nationality, membership of a particular social group, or
Cambridge Books
political opinion” – illustrates illegitimate reasons for inflicting serious

harm. It reflects a distinctively liberal conception of legitimacy: one that
protects pluralism by promoting the values of equality, religious tolera-
tion, and political accountability, while still leaving states with wide
latitude to enact policies that respond to local needs or reflect local
values and culture. In light of these values, the bases for harm identified
by the nexus clause are manifestly at odds with the usual presumption
that the state acts for the benefit of its citizenry.
The specification of race and nationality as illegitimate reasons for
harm reflect the self-evident proposition that a state that targets sectors of
its population for harm on the basis of immutable, ascriptive character-
istics cannot possibly be said to act on their behalf. The same is true of
harm inflicted on account of religion. Although a state need not be
neutral with respect to religion, it must nonetheless tolerate minority
faiths by allowing their members to exercise their beliefs freely. The
same could also be said about harm inflicted on account of other ascrip-
tive characteristics not explicitly delineated in the nexus clause, such
as ethnicity, disability, kinship, sexual orientation, and gender.
“Membership in a social group” should be read to encompass such
attributes. That reading is consistent with the gloss given by current US
law: a “social group” is understood to refer to any group sharing “a
common, immutable characteristic,” including “an innate one such as
sex, color, or kinship ties,” as well as “a shared past experience such as
former military leadership or land ownership … [W]hatever the common
characteristic that defines the group, it must be one that the members of
the group either cannot change, or should not be required to change
because it is fundamental to their individual identities or consciences.”24
Political opinion is an impermissible basis for harm because the free-
dom to express grievances is necessary if leaders are to be held accoun-
table for official misconduct or errors in judgment. A regime need not be
democratic to enjoy freedom from external interference, but it must
allow its citizens to assemble for peaceful protest, to publish dissenting
opinions in the media, and to hold diverse viewpoints. A regime unwill-
ing to permit criticism does not take seriously its responsibility to act on
behalf of its citizenry. As Immanuel Kant put it long ago, “the citizen
must … be entitled to make public his opinion on whatever of the ruler’s
measures seem to him to constitute an injustice against the common-
wealth … Freedom of the pen is the only safeguard of the rights of the
people …”25
Cambridge Books
The idea that political opinion is an illegitimate basis for inflicting

harm can explain the doctrine of imputed political opinion, which holds
that people can be persecuted on account of a political opinion that they
do not actually hold but that is imputed to them by a persecutor. “What
justifies refugee protection” in such a case, explains Daniel Steinbock,
“is that the political opinion that the persecutors attribute to the victim
is an … improper[ ] ground for punishment,” regardless of whether the
victim actually espouses that opinion.26
In one important way, the nexus clause is too narrow. In focusing on
various personal characteristics as illegitimate bases for harm, it fails to
capture what I shall call liberalism’s anti-brutality norm: the idea,
reflected in international human rights law, that some harms are so
serious that their infliction for any reason is inconsistent with the pre-
sumption of legitimacy. For example, customary international law recog-
nizes as peremptory the rights to be free from genocide, slavery, murder
or disappearance of individuals, torture, prolonged arbitrary detention,
and apartheid.27 These harms are always outside the boundaries of
permissible state action, no matter what the reason for inflicting them.
People may reasonably disagree about the scope or interpretation of
these peremptory human rights – for example, whether capital punish-
ment is simply judicial murder or can sometimes be legitimate – and they
may also disagree about whether other rights should be added to this
list. For example, the anti-brutality norm may also bar serious viola-
tions of international humanitarian law, so that states may never target
civilians or carry out military operations that expose them to harm that
is disproportionate to the military objective. But bracketing these dis-
putes about interpretation and scope, the notion that there are some
absolute constraints on governmental action is well accepted.
“Subsidiary protection” in the European Union can be understood, in
part, as implementing the anti-brutality norm. That form of protection
effectively extends asylum to people who would, if returned to their country
of origin, face the death penalty, torture, or other inhuman or degrading
punishment.28 In the view of the EU, all three categories of treatment can
never be justified by a state’s legitimate interest in punishment. They are in
every case beyond the bounds of legitimate state authority.
Recognizing an anti-brutality norm would require a change to US
law. Currently, applicants must show that they have been persecuted
because they possess, or have had imputed to them, one of the char-
acteristics listed in the nexus clause: race, religion, nationality, social
Cambridge Books
group membership, or political opinion. The list of grounds given by the

nexus clause is an exclusive one. This has had unhappy consequences in
many cases, most notoriously those involving police investigations.
Matter of R – is representative of such cases. A 21-year-old Sikh man
was visited by Sikh militant separatists at his home in Punjab and was
asked for money in support of their cause. He declined to give them
money because he advocated separatism only through non-violent
means. He was subsequently arrested by the Indian police and was
brutally tortured while being interrogated about the militants’ visit.
He was eventually released without charges, and upon his arrival
home, was confronted again by the militants, who beat him and threa-
tened to kill him and his family unless he joined them. Fearing harm
from both the militants and the police, R fled to the United States.29
The US Board of Immigration Appeals (BIA) rejected his asylum
claim on the grounds that neither:
the Sikh militants [nor] the police who confronted the applicant sought to punish
him on account of one of the grounds enumerated in the [nexus clause] … [T]he
Sikh militants were seeking operating resources from the applicant in the form of
material assistance and manpower … Similarly, there is no indication that the
police actions against the applicant extended beyond the investigation of and
reaction against those thought – rightly or wrongly – to be militants seeking the
violent overthrow of the government … While the applicant states he was
subjected to police brutality, which we certainly do not countenance, the record
reflects that the purpose of the mistreatment was to extract information about
Sikh militants, rather than to persecute the applicant “because” of his political
opinions or the mere fact that he was a Sikh.30
In other words, R failed to qualify for asylum because the police

tortured him to gain information they believed he possessed; the mili-
tants beat him in order to recruit additional manpower; and these are
not protected grounds under the nexus clause. That approach to the
nexus clause has rightly struck many observers as troubling.31
One could argue that the BIA mischaracterized the police’s reasons
for harming R, and that the torture was at least in part on account of
political opinions they imputed to him. But even assuming that the BIA’s
analysis of the facts was correct, and that the police’s sole reason for
torturing R was to gain information that they believed he possessed, R
should have been granted asylum. Customary international law would
certainly regard the police torture of R as having violated a peremptory
Cambridge Books
human rights norm, reflecting the judgment that such treatment can
never be justified for any reason. Extending asylum eligibility to people
like R follows straightforwardly from a political conception of asylum.*
It is true that people like R are often eligible for protection against
deportation under the Convention Against Torture (CAT). For exam-
ple, in the United States, an alien cannot be removed if he “is more likely
than not to be tortured in the country of removal.”32 But that is a more
stringent standard than in asylum proceedings: to benefit from CAT,
applicants must prove that they are “more likely than not” to be
tortured, while asylum applicants must prove only a “well-founded
fear” of persecution. A “well-founded fear” may exist even if there is
a 10 percent probability of harm.33 Moreover, protection under CAT is
a much less desirable status than asylum. In the United States, it does not
entitle one to family reunification or place one on the path to permanent
residence and citizenship. Under the EU’s directive providing for sub-
sidiary protection, recipients are guaranteed only a renewable one-year
grant of protection, as opposed to a renewable three-year grant of
protection for those qualifying as Convention refugees.34
One approach to incorporating an anti-brutality norm into current law
is to add to the Convention refugee definition another set of grounds on
which an asylum claim could be based. This is roughly what the EU has
done with the device of subsidiary protection (although the EU continues
to distinguish between Convention refugees and recipients of subsidiary
protection in terms of the relief given). Another approach would be to
adopt a statutory presumption that victims of torture and other “brutal”
treatment have suffered persecution on account of a Convention reason.
The US Board of Immigration Appeals has suggested in passing that
“there may be situations in which the severity of the violations of the
Geneva Convention may support an inference that the abuse is grounded
in one of the protected grounds under the asylum law.”35 Such an
inference could be presumed for the violation of peremptory human
rights norms, including torture.36 As we shall see below, the United
States has, in effect, adopted such a presumption for victims of China’s
* One may object that granting asylum to those who suffer torture or abuse during a
legitimate police investigation would mean granting refuge to terrorists and
criminals who have experienced mistreatment by their states of origin. But there is
no such danger. Separate rules under both international and domestic law exclude
such persons from asylum. I further discuss these “bar statutes” below.
Cambridge Books
coercive sterilization program, and a similar statutory presumption was

proposed for victims of female genital mutilation.
Constraining adjudicators
Because people can reasonably disagree about whether particular rea-
sons are a legitimate basis for harm and about whether certain treat-
ment is proscribed by a peremptory human rights norm, adjudicators
will inevitably need to rely on their own judgment as to the bounds of
legitimate state authority. Some might raise the specter of adjudicators
deciding cases based on nothing other than their own subjective moral
assessments. But, in fact, a variety of mechanisms meaningfully con-
strains adjudicators and ensures that their decisions reflect collective
judgments about legitimacy.
First, adjudicators are constrained by statute, which – as just dis-
cussed – outlines the types of reasons that should be regarded as
illegitimate bases for harm. Any determination needs to remain
anchored to the statute. Second, adjudicators are constrained by pre-
cedent. Over time, through the accretion of decisions, interpretive gaps
are narrowed and a consensus will emerge on the types of harms
regarded as sufficiently serious to constitute persecution and on the
reasons for harm regarded as illegitimate. At the same time, this con-
sensus is dynamic, and our notion of what constitutes persecution can
shift with changing moral sensibilities. For example, courts have been
increasingly open to the idea that sexual orientation and gender should
be regarded as “social groups” under the asylum statute, so that harm
inflicted on their account can give rise to asylum eligibility.37
Third, normative disagreements can be at least partly resolved
at the policy-making level through the use of administrative guidelines
and legislative presumptions. For example, the Immigration and
Naturalization Service (INS) Gender Guidelines, designed to create
“uniformity and consistency” in gender-related claims, alerted
American adjudicators to a variety of considerations relevant to such
claims. The BIA, whose members are appointed by the Attorney
General, issues decisions that set binding agency policy. In addition,
the legislature always remains free to step in and overrule a policy
judgment made by the BIA or the courts that it dislikes.
The law concerning Chinese coercive population control policy pro-
vides a good example. The BIA had held that the Chinese program was
Cambridge Books
not persecutory because, as a law of general application, it did not target

people for harm on account of any of the characteristics listed in the
nexus clause. Congress responded by enacting a statute deeming invo-
luntary abortion and involuntary sterilization to be persecution “on
account of political opinion” in all cases – even those lacking any evi-
dence of political opinion. Effectively, the statute treated coercive popu-
lation control as falling within the anti-brutality norm: as beyond the
scope of legitimate government action no matter what the reason for it.38
The role for human rights

In Chapter 2, I suggested that asylum can be linked to a human rights-
oriented foreign policy, and above I suggested that international
human rights law can offer guidance in determining the scope of the
anti-brutality principle. Respect for core human rights, that is, those
rights recognized by customary international law as peremptory, is a
necessary condition for legitimate state conduct, and so treating the
violation of those basic rights as giving rise to asylum eligibility is
consistent with the political conception of asylum. But, under the
approach I have suggested, legitimacy should remain the touchstone
for defining “persecution,” not human rights.
That approach runs counter to the main trend among refugee scho-
lars in recent years, which has been to couple asylum policy to interna-
tional human rights law, so that the violation of any of the human rights
listed in the Universal Declaration of Human Rights (UDHR) and
codified in either the International Covenant on Civil and Political
Rights (ICCPR) or the International Covenant on Economic, Social,
and Cultural Rights (ICESCR) would constitute persecution. This
“human rights approach” (as I shall call it) was pioneered by Professor
James Hathaway, has been embraced by other prominent academics and
activists, and has been endorsed by courts in Canada, Britain, Australia,
and New Zealand.39
The rationale is that the human rights listed in the Covenants repre-
sent “the minimum duty owed by a state to its nationals,” which “[a]ll
states are bound to respect as a minimum condition of legitimacy.”40
When a state “ignores or is unable to respond to legitimate expectations
as defined in international human rights law [by failing] to comply with
its most basic duty,” citizens of that state should have the “prospect of
legitimate disengagement from that community in favour of surrogate
Cambridge Books
protection elsewhere.”41 Thus, Hathaway concludes, “persecution may

be defined as the sustained or systemic violation of human rights
demonstrative of a failure of state protection.”42
Human rights fall into four categories. In the first are rights from
which no derogation is ever permitted, including not only the peremp-
tory human rights norms covered by the anti-brutality principle, but
also a right against ex post facto criminal punishment and rights to
legal personhood and freedom of thought, conscience, and religion.
The second category includes rights from which derogation is per-
mitted in times of public emergency, including freedom from arbitrary
detention, the right to a fair trial in criminal proceedings, privacy
rights, rights to political participation, the right to join trade unions,
and freedom of opinion, expression, assembly, and association. The
third category consists of economic and social rights, such as rights to
food, clothing, housing, health care, and social security; states are
obligated to “take steps” to achieve these rights to the extent that
their resources allow. In the fourth group are rights listed in the
UDHR, but not codified in either the ICCPR or ICESCR, such as the
right to be free from arbitrary deprivation of property. Hathaway
concludes that the violation of rights in this fourth category does not
constitute “persecution.”43
In effect, the human rights approach expands the scope of the anti-
brutality principle from the core of basic human rights recognized as
binding under customary international law to the full roster of rights
listed in international instruments. One consequence of this approach
would be radically to de-emphasize the nexus clause. Rather than
determining whether harm is inflicted for an illicit reason, the human
rights approach identifies a large class of harms that cannot be inflicted
for any reason.44
The human rights approach is helpful in giving additional texture to
the concept of persecution and thereby promoting consistency of
decision-making. But “persecution” should not be defined solely
through the prism of human rights. Instead, legitimacy must remain
the primary conceptual framework. This is so for several reasons.
Overinclusion and underinclusion

First, the human rights approach reifies a list of human rights that was
the product of both political compromise and aspirational thinking.
Cambridge Books
The human rights codified in the ICCPR and ICESCR may in the
aggregate be good proxies for legitimacy, but they are not determinative
of legitimacy. The list is both overinclusive and underinclusive. As an
example of overinclusion, consider the right “[t]o vote and to be elected
at genuine periodic elections,”45 derogable only in cases of national
emergency. One might find it implausible that every citizen of a benign
and decent monarchy has been persecuted. Perhaps democracy is the
most just method of governance; but a decent monarchy is not liable to
external intervention or deserving of international condemnation
merely due to the nature of its political system. To award asylum to
citizens of a monarchy on such a basis is at odds with the political
conception. Similarly, one might find implausible the notion that a
wealthy state’s failure to provide social security46 violates the “mini-
mum conditions” of its legitimacy, thereby permitting poor would-be
pensioners to claim asylum elsewhere. Rectifying this kind of distribu-
tive injustice is not what asylum is for.
In light of examples such as these, it is unsurprising that the
Commonwealth courts adopting the human rights approach emphasize
that eligibility for asylum is limited to applicants who can show a well-
founded fear of serious harm in addition to a violation of human rights.
“ ‘Being persecuted,’ ” they have held, “is the construct of two separate
but essential elements, namely risk of serious harm and a failure of state
protection” of human rights.47 The requirement that human rights
violations be accompanied by serious harm dramatically curtails the
scope of the human rights approach. Some human rights violations
inherently involve serious harm, most notably the peremptory human
rights norms encompassed by the anti-brutality principle, like slavery,
torture, murder, and prolonged arbitrary detention.
However, for most rights listed in the ICCPR and ICESCR, their
violation does not inherently constitute serious harm; and in the case
of many, it is hard to see how their violation could ever constitute
serious harm. In the last category fall rights to political participation
and many economic, social, and cultural rights, such as the rights of
trade unions, the right to maternal leave, the right to reasonable limita-
tion of working hours, and the right to equal pay for equal work. In
sum, the requirement that human rights violations must be accompa-
nied by serious harm in order to constitute persecution effectively
acknowledges that the roster of human rights in the ICCPR and
ICESCR is overinclusive for purposes of asylum policy, and that the
Cambridge Books
violation of such human rights does not necessarily violate the mini-
mum conditions of a state’s legitimacy.
At the same time, the list of human rights in the ICCPR and ICESCR
is also underinclusive from the perspective of legitimacy. Limiting asy-
lum to those who have been denied those rights alone would leave
unprotected several categories of people who should be eligible for
asylum. For example, the “right to own and be free from arbitrary
deprivation of property”48 is not included in either the ICCPR or the
ICESCR and, therefore, on Hathaway’s scheme, its violation could not
constitute persecution. But American courts rightly agree that, in some
circumstances, “confiscation of property” can “cross the line from
harassment to persecution.”49
Punishment for “absolute political offenses” – such as a coup or other
criminal act directed against the existence of a state or against a head of
state50 – offers another example of underinclusion. Punishing coup-
plotters or rebels does not violate their human rights. The ICCPR
recognizes a right to political expression, but it is subject to a limitation
for “the protection of national security or of public order.”51 Rebellion
falls outside the scope of this right on two counts: it is action rather than
expression; and even if it were regarded as expression, it would certainly
come within the exception for national security and public order. This
result is unsurprising, given that the human rights conventions were the
result of negotiation among governments which presumably were
unwilling to forfeit their power to punish traitors and insurrectionists.
But given asylum’s historical role of sheltering revolutionaries from
punishment – in particular, revolutionaries who have fought against
tyrannical regimes – the conclusion that insurrectionists are ineligible
for asylum is quite unsatisfying. Later on in this chapter, I shall have
more to say about insurrectionists and their eligibility for asylum.
Human rights are not self-defining

The second reason that legitimacy, rather than human rights, must be
the touchstone for asylum is that human rights are not self-defining.
They require interpretation. What does it mean to be “treated with
humanity and respect for the inherent dignity of the human person”?
What constitutes “arbitrary” arrest or detention? How extensive is the
right to “freedom of expression” or “freedom of association” or “free-
dom of conscience”? Similar interpretive problems arise concerning
Cambridge Books
socio-economic rights, which are violated only when a state fails to

“take steps” to secure them to the extent that their resources allow. But
what constitutes an adequate “step”? Resources are scarce; how much
latitude should states have in allocating them as they deem appropriate?
In sum, we must look beyond the ICCPR and ICESCR to fill out the
content of human rights. In defining their scope, we are necessarily
thrown back on an underlying theory of legitimacy. This should be
unsurprising. After all, the list of human rights in the ICCPR and
ICESCR did not drop from the sky, but instead were themselves distilled
from widely shared notions of legitimacy.
To take just one of many possible examples, how should one regard
laws enforcing Islamic dress, for example, Iran’s law requiring women
to wear chadors, or hair coverings, when in public? Saideh Hassib-
Tehrani, a native of Iran, entered the United States in 1984 and applied
for asylum shortly thereafter. She claimed that she had been questioned
and briefly detained on a couple of occasions for having violated
various Islamic morals regulations. One time, she had attended a
party at a male friend’s house and had seen him in a bathing suit. She
was later detained by the authorities for several hours with other female
guests and was instructed that “being present with a man in a bathing
suit was incorrect.” Another time, she was stopped on the street by four
government officials because she “had a few pieces of hair hanging out
of her [chador] by mistake,” was forced into their car at gunpoint,
admonished “not to appear on the street like that again,” and returned
home. Hassib-Tehrani testified that she had been so greatly distressed
by these events that she became ill, missed several months of work, and
fled Iran.52 Her asylum application argued that the enforcement of the
chador law constituted persecution.
How would her case be analyzed from the standpoint of the human
rights approach? One possibility would be to invoke the “right to free-
dom of expression.”53 In fact, that label ill fits Hassib-Tehrani’s case. She
objected to the chador law, but she did not express her objections pub-
licly. The one time she was detained for having violated the law, it was
because she had left some hair exposed “by mistake.” Nor did she testify
that, if she returned to Iran, she would openly defy the law in protest. But,
counterfactually, we can imagine an activist who would openly violate
the law upon returning to Iran to express her opposition to it.
The right to freedom of expression can be limited, says the ICCPR,
“[f]or the protection of national security or of public order … or of
Cambridge Books
public health or morals.”54 One could view the chador law as a morals
regulation akin to Western laws forbidding public indecency. Morals
regulations of this sort enforce culturally-based views about what
should remain hidden from public scrutiny. Such views differ from
society to society – even within the West. For instance, German
women visiting American beaches are sometimes surprised to discover
that removing their tops is against the law.55 Few would think that
nudists are persecuted when they are punished for venturing outside
without any clothes – even if nudism has some expressive significance.
Governments may legitimately enforce general criminal prohibitions,
even against political activists who violate them in order to express their
opposition.56 Otherwise, as Steinbock has noted, a bank robber could
claim that his robbery expressed the political opinion that “capital is
theft” and that his punishment would, therefore, be persecutory.57
A better argument is that the chador law amounts to sex discrimina-
tion – indeed, it is just one facet of a social system that severely dis-
criminates on the basis of sex – and that the “public morals” account
neglects this social reality. Requiring women to wear a chador when in
public is a means of controlling them. And the ICCPR guarantees
“equal protection of the law” and enumerates a right against “any
discrimination” on the basis of sex.58
But the “public morals” account has some force in the face of this
objection as well. Even in the West, stricter standards of decency apply
to women than to men; yet public decency laws are nonetheless
regarded as laws of general application. Iranian laws regulating public
decency are certainly more invasive of personal freedom than Western
ones, and the gap between what is forbidden for women and for men is
greater, but perhaps that should be regarded as a difference in degree
rather than in kind. Along these lines, the court concluded in Hassib-
Tehrani’s case that she “merely … faces a possibility of prosecution for
an act deemed criminal in Iranian society, which is made applicable to
all [women] in that country,” and that “prosecution for general crimes”
should be distinguished from persecution.59
Whether the chador law should be regarded as a public morals
regulation or as sex discrimination cannot be answered from the text
of the ICCPR alone. Rather, any answer will turn on an underlying
conception of legitimacy that gives texture to those rights. Surely there
are limits to what can be imposed on women in the name of public
morals. Imagine, for example, a society which regarded painful and
Cambridge Books
disfiguring foot-binding as essential to decency. A law requiring women

to undergo foot-binding would rightly be regarded as persecutory. Or,
less hypothetically, imagine a society which required women, either as a
matter of law or as a legally condoned cultural practice, to undergo the
painful and disfiguring removal of part of their genitalia.
The court’s depiction of the chador law in Hassib-Tehrani’s case – as
a law of general application – is therefore parasitic on an implicit
judgment that Iran is within its latitude to impose such burdens on
women in pursuit of a communal interest in regulating morals. On
that view, the chador law is more like Western laws proscribing topless
sunbathing than like foot-binding or female genital mutilation. The
court’s conclusion also reflects the likelihood that many Iranian
women support the chador law.
Where exactly one should draw the line between a ban on topless
sunbathing and foot-binding – at chadors? at burkas? somewhere else? –
is not self-evident. We might conclude that while a chador law passes
muster, a law requiring women to wear burkas – which interfere with
vision – does not, and is better described as illegitimate sex discrimination
than as a legitimate regulation of public morals. The answer will inevitably
draw upon one’s own culturally influenced intuitions as to when the
burdens imposed upon women in the name of public decency are so severe
that the presumption of legitimacy is overcome. Nothing in the ICCPR
compels an answer one way or another, and in cases like Hassib-Tehrani’s,
reasonable people may well have very different intuitions.60
Human rights in conflict

The third reason that legitimacy rather than human rights must be the
touchstone in defining persecution is that human rights can come into
conflict. As Michael Ignatieff has noted, politics is rife with tragic
choices.61 The case Matter of Chang, concerning Chinese coercive popula-
tion control measures, offers a useful example of such a conflict.62 Chang
claimed that he was afraid to return to China because he believed that he
would be subjected to forced sterilization for having had a second child
with his wife, in contravention of China’s “one couple, one child” policy.63
The couple had already been forced to leave their commune after they were
given no work to do, and both Chang and his wife had been ordered to
report to a sterilization clinic. His wife had been able to postpone her visit
because of an illness, but Chang decided to flee the country.64
Cambridge Books
The BIA rejected Chang’s application. The population control law, it

reasoned, was one of general application; in the absence of any evidence
that the policy was “a subterfuge for some other persecutive purpose,”
Chang was not eligible for asylum.65 The BIA’s decision was heavily
criticized by human rights activists, who endorse a Canadian decision as
a template for such cases.66 The Canadian court described forced ster-
ilization as a “serious and totally unacceptable violation of [one’s] secur-
ity as a person. Forced sterilization … is such an extreme violation of their
basic human rights as to be persecutory, even though this was thought to
advance the modernization of China.” The court concluded, “Brutality in
furtherance of a legitimate end is still brutality.”67
While the Canadian court may have reached the right outcome, its
analysis skates over the complexity of the situation. China adopted
coercive population control measures because it judged them to be the
lesser of two very unpleasant options. As the BIA explained:
China has adopted a policy whose stated objective is to discourage births

through economic incentives, economic sanctions, peer pressure, education,
availability of sterilization and other birth control measures, and use of
propaganda. Chinese policymakers are faced with the difficulty of providing
for China’s vast population in good years and in bad. The Government is
concerned not only with the ability of its citizens to survive, but also with their
housing, education, medical services, and the other benefits of life that persons
in many other societies take for granted. For China to fail to take steps to
prevent births might well mean that many millions of people would be
condemned to, at best, the most marginal existence.68
On the Board’s telling, China believed that it faced a stark and tragic
choice. Either it adopted extreme measures, or it would, before long,
be unable to satisfy the basic subsistence needs of its population.69
China’s solution may have been the wrong one, but human rights lie
on both sides of the balance. An analysis of Chang’s case must look
beyond the roster of human rights to an underlying conception of
legitimacy that explains why some rights should be given priority over
others.
***
To sum up: in arguing against the human rights approach to defining
persecution, I have sounded a number of variations on a theme. In
deciding whether a government measure constitutes persecution, the
Cambridge Books
roster of human rights given by the international covenants is an unreli-

able guide, and one must often fall back on an underlying theory that
describes the limits of legitimate government authority. The roster of
human rights is both over and underinclusive as a proxy for legitimacy;
human rights themselves are not self-defining but require interpretation
in light of underlying principles; and human rights often conflict,
requiring a decision-maker to rely on a theory that explains why some
rights should be prioritized over others. For these reasons, legitimacy
rather than human rights must remain the touchstone in determining
what constitutes “persecution.”
Hard cases
In many, indeed, most cases, it will be clear whether a state’s reason for
inflicting harm should be presumed legitimate or should be regarded as
inconsistent with that presumption. But other cases are less obvious.
Below, I discuss a few of these hard cases.
The case of rebels and other political offenders

How should asylum law treat coup plotters, insurrectionists, and others
who commit quintessentially political offenses, directed at the security
of the state and its leaders? Should their punishment be regarded as
legitimate prosecution for treason or instead as political persecution?
Consider two straightforward, but mistaken, answers to the question.
The first is that governments always have the legitimate right to punish
rebellion. The case Matter of Maldonado-Cruz offers an example of this
erroneous approach. The applicant in that case was a Salvadoran who
had been kidnapped by guerrillas in 1983 and impressed into their army.
A few days later, he managed to escape and flee to the United States. He
was afraid that, were he returned to El Salvador, he would be killed by the
government for being a guerrilla. The BIA denied his claim to asylum,
reasoning that because “the Government of El Salvador is … a duly
constituted and functioning government of that country, it has the
internationally-recognized right to protect itself against the guerrillas
who seek to overthrow it.”70
That analysis confused a de facto exercise of power with a legitimate
exercise of power.71 A government can be “duly constituted and func-
tioning” but still conduct itself in ways that make it liable to justified
Cambridge Books
resistance and overthrow. Indeed, a limited right to revolution lies at the

core of the liberal tradition. Sovereign authority is not absolute, and one
of asylum’s core historic purposes was to facilitate the right to revolt by
sheltering revolutionaries who had waged unsuccessful rebellions
against autocratic regimes.72
The second straightforward but mistaken answer is to say that,
because all rebellions are political, punishment for rebellion should
always be regarded as harm inflicted on account of political opinion,
giving rise to eligibility for asylum. But condemnation is not always an
appropriate response to the suppression of rebellion. Some govern-
ments do have a “legitimate right” to defend themselves against rebels.
What we need are criteria for identifying such governments.
One possibility is that democracies may legitimately punish rebels,
but non-democracies may not. Rebellion, the argument goes, is a legit-
imate vehicle for political expression when a country lacks “elections to
governing organs.” In such countries, “a coup is the only means
through which a change in the political regime can be effected … In
this context, [a coup plotter’s] political expression is embodied in his
political act.”73 Applying this criterion, an American court in Dwomoh
v. Sava ruled that a Ghanaian coup plotter was eligible for asylum.
The democracy–nondemocracy distinction reflects a misguided
notion of international legitimacy. For example, a decent monarchy –
one that respects the rule of law, protects liberty, and is willing to accept
criticism and advice from the populace – should enjoy freedom from
external interference and international condemnation. States should
have latitude to choose institutional arrangements that resonate with
their particular traditions and customs. Conversely, an illiberal democ-
racy – one that persecutes its minorities – should not enjoy any such
entitlement.
A better approach would look to whether the state tolerates peaceful
expression of dissenting viewpoints. If so, then it may legitimately
punish those dissenters who choose to use force. Applying this principle,
another American court denied asylum to a Filipino military officer
charged with attempting a coup against Aquino in 1987. In ruling
that his punishment constituted legitimate prosecution rather than poli-
tical persecution, the court emphasized that “diverse political views are
tolerated in the Philippines.” It concluded that “prosecution for parti-
cipation in a coup does not constitute persecution on account of poli-
tical opinion when peaceful means of protest are available for which the
Cambridge Books
alien would not face punishment.”74 A coup is defensible only when

normal avenues of political protest are blocked.
***
Even if rebels do have a legitimate right to rebellion, they must still
conduct their rebellion using legitimate tactics. The distinction familiar
from the law of war – between jus ad bellum and jus in bello – has
application in this context. Recall from Chapter 1 the case of the
“wretch Fieschi,” who killed eighteen bystanders in his unsuccessful
attempt on the life of King Louis-Philippe of France in 1835.75 Even if
Fieschi had a legitimate right to rebel, he could nonetheless be legiti-
mately punished by France: not for the mere act of rebellion, but instead
for the carnage he inflicted in the process. Asylum should not protect
him from such punishment. Accordingly, people who commit serious
non-political crimes, even if those crimes were politically motivated, are
ineligible for asylum.76
At the same time, however, a state’s legitimate right to investigate and
punish people like Fieschi does not give it a blank check to use extreme
tactics. Consider Matter of R –, the case involving the Sikh man who
was tortured for information regarding separatist Sikh militants. It is
true, as Board member Heilman argued in concurrence, that “the
political program of the Sikh extremists” – “the violent separation of
Punjab from India through a program of murder and terrorism” – could
be punished legitimately.77 The Indian state permitted peaceful dissent
and, in fact, engaged in negotiations with non-violent Sikh activists; and
moreover, the tactics of the Sikh extremists were in the same class as
those of Fieschi. But R himself was not a militant, and the inhuman
treatment to which he was subjected as a non-belligerent “was dispro-
portionate to any measured government reaction that could be expected
as a response to anti-government activity.”78 A government’s general
right to repress uprisings does not justify all action taken toward that
end. The anti-brutality principle constrains what a state may legiti-
mately do in self-defense.
But what happens when an illegitimate rebellion is met by illegitimate
investigatory tactics? Suppose, for example, that one of the Sikh mili-
tants – clearly deserving of punishment not only for rebelling against a
state that permits lawful dissent but also for committing atrocities in the
process – is likely to be brutally tortured if he is returned to India.
Should he be granted asylum? It is true that he faces treatment (torture)
Cambridge Books
deserving of condemnation. But are his own deeds so odious that he

forfeits any claim to asylum’s benefits? A similar question arises con-
cerning terrorists who are justly sought for punishment by their states,
but who are likely to be tortured if returned there. Should the policy
concerns that drive asylum, that is, the promotion of rule of law and
respect for basic human rights, give way to other policy concerns, like
ensuring that particularly dangerous criminals are punished, even if
excessively or inhumanely so?
The political conception of asylum points both ways. On the one
hand, asylum is intended (as Grotius wrote) “for the benefit of those
who suffer from undeserved enmity, not those who have done some-
thing that is injurious to human society or to other men.”79 This
suggests that asylum should not be available to those deserving of
punishment for their deeds – even if the punishment is excessive. That
view echoes not only Grotius, but also Vattel, who argued that crim-
inals “who by the character and frequency of their crimes are a menace
to public security everywhere”80 ought to be extradited and punished,
not granted asylum. On the other hand, excessive or cruel punishment is
never deserved; and a state whose justice system endorses such punish-
ment is deserving of the condemnation that asylum expresses. Recall the
English tradition, which extended asylum to any fugitive from a des-
potic state – including not only innocent victims but also (as Jefferson
put it) “the most atrocious offenders … who have been able to get
there.”81
Although theory pulls in both directions, as a practical matter states
cannot be expected to shelter asylum seekers who pose a danger to the
community of refuge. The American regulations governing both asylum
and CAT reflect this view. Neither form of relief is available to aliens
who have “ordered, incited, assisted, or otherwise participated” in the
persecution of others; who have committed a “serious nonpolitical
crime outside the United States”; or who pose a “danger to the security
of the United States”82; or, in the case of asylum, who have “engaged in
terrorist activity.”83 Other states have similar provisions, as does the
UN Convention.84
Such exclusions, however, should be applied carefully so as not to
sweep in the innocent. For example, a person has not “participated in
the persecution of others” merely because his actions had the effect of
furthering persecution. The term “persecution” implies a culpable men-
tal state on the part of the persecutor. Thus, the “bus driver who
Cambridge Books
unwittingly ferries a killer to the site of a massacre can hardly be labeled

a ‘persecutor,’ even if the objective effect of his actions was to aid the
killer’s secret plans.”85 Nor has one engaged in persecution if one’s
actions were carried out under duress.86
In this respect, American law is problematic in its expansive defini-
tion of “engaging in terrorist activity,” which includes committing any
“act that the actor knows, or reasonably should know, affords material
support” to a terrorist organization or to any individual who the actor
has reason to know has committed or plans to commit terrorist activ-
ity.87 “Material support,” in turn, is defined broadly, and includes such
acts as providing food, shelter, or transportation. There is no exception
for acts undertaken in duress. The consequence has been the denial of
asylum to a nurse from Colombia who was kidnapped and forced to
provide medical treatment to members of the FARC, a terrorist group; a
fisherman from Sri Lanka who was abducted by the Tamil Tigers and
forced to pay his own ransom; and a journalist from Nepal who was
beaten and forced to pay money to Maoist terrorists.88 The material
support bar has also been used as the basis for rejecting asylum seekers
from Iraq who had paid ransoms to free family members kidnapped by
Shiite or Sunni militias. The Department of Homeland Security has sub-
sequently determined that material support provided under duress should
not disqualify an asylum seeker – but only if the material support was
provided to terrorist groups not specifically designated as such by the
Secretary of State.89 This concession will do little to help victims of groups
like the FARC, the Tamil Tigers, and al-Qaeda in Iraq, all of which have
been designated “foreign terrorist organizations” by the State Department.
Laws of general application

Laws that apply generally to the entire population enjoy a very strong
presumption of legitimacy; they are assumed to reflect the will of the
community that is bound by them. There are narrow exceptions to this
rule. For example, I have argued that particularly brutal policies – like
torture or other violations of peremptory human rights – are illegitimate
even if they are generally applied. These norms form side constraints on
states’ freedom of action. General laws can also sometimes be a sub-
terfuge for discrimination; imagine, for example, a ban on male circum-
cision in a society generally repressive toward Jews. One might infer
that the ban, though phrased in general terms, is in substance a measure
Cambridge Books
directed at Jews. One can also imagine a general law that is enforced
discriminatorily. But notwithstanding these narrow exceptions, the
usual rule is that communal self-determination is advanced by out-
siders’ respect for laws of general application.
Not infrequently, general laws have an unintended adverse impact –
sometimes quite a severe one – on minority populations. For example, a
law regulating animal slaughter might interfere with the sacrificial
practices of a religious minority; similarly, a law proscribing psychede-
lic drugs may interfere with a religion’s ritual practices. How should
asylum law address situations like these? To avoid persecuting religious
minorities, must a state accommodate their practices by exempting
them from laws of general application?
The case of the Canas brothers, Jehovah’s Witnesses from El
Salvador, provides a concrete example. The Canas brothers’ religion
forbade them from participating in military service. But El Salvador did
not exempt conscientious objectors from its mandatory military draft,
nor did it make available to conscientious objectors any other form of
national service. Violators of the draft faced between six months and
fifteen years in prison. The BIA denied the Canas brothers asylum on the
ground that the conscription law “applies equally to all Salvadorans.”90
In one sense, that was true: the draft law was generally applicable. But
Jehovah’s Witnesses were differently situated than other Salvadoran
citizens who had no conscientious objection to military service; the
Jehovah’s Witnesses needed either to violate their religious beliefs by
serving or face jail time for draft evasion. That was not a choice that
other Salvadorans needed to face. In that sense, the law did not apply
equally to all Salvadorans: it burdened some citizens much more
severely than others on account of their religion. Did El Salvador’s
failure to exempt the Canas brothers constitute religious persecution?
Or was El Salvador entitled “simply [to] insist[ ] on universal military
service for all citizens” without any such exemption?91
One approach would be to say that the presumption of legitimacy
that attaches to generally applicable laws can be overcome only by
evidence showing that the law is a pretext to inflict harm for an illegi-
timate reason. The burden of producing that evidence would lie on the
asylum applicant. For example, a general law that disparately impacts a
religious minority might be regarded as pretextual if the religious
minority otherwise faces discrimination or oppression, or if there are
no possible legitimate reasons for the general law. Under this approach,
Cambridge Books
the Canas brothers would certainly lose. A general conscription law is

not a pretext for religious persecution; rather, the state needs to raise
an army.
A second approach holds that a state must exempt disparately
affected minorities from a general law unless it has some legitimate
reason for not doing so. The burden of producing such a reason
would lie on the immigration service. This approach is somewhat
more favorable to asylum seekers, but the Canas brothers would lose
under it as well. There are several good reasons why a state may decline
to exempt Jehovah’s Witnesses from conscription laws: an exemption
would open a loophole that could be exploited by those simply unwill-
ing to fight; it can be costly to administer especially at a time of national
emergency; and it is at odds with the republican principle that all those
who enjoy the benefits of citizenship should also bear its burdens.
A final approach, significantly more protective of religious minorities,
would require states to exempt disparately affected minorities from gen-
eral laws whenever the affected group is “unduly burdened” by the law.
The Ninth Circuit Court of Appeals initially took roughly this approach
in the Canas brothers’ case.92 It would require courts to engage in
substantially more fact-finding than either of the first two approaches.
An adjudicator would need to consider, for example, how costly an
exemption would be to administer and the degree to which an exemption
would subvert the state’s regulatory scheme. Those interests would then
need to be weighed against those of the religious minority’s. On that side
of the scale, one would need to consider how central a given practice is –
for example, non-violence or psychedelic drug use – to the religion.
When the problem is viewed through the prism of legitimacy, one is
guided toward the first or second approach rather than the third. The
concept of legitimacy affords states wide latitude to make policy deci-
sions, and even unjust policies are not illegitimate if they are enacted in
good faith. Here, there are several good faith reasons for a generally
applicable draft law that makes no exception for conscientious objec-
tors. Unless there is countervailing evidence to support an inference of
animus, the Jehovah’s Witnesses should lose. And, indeed, in similar
cases, many American courts have so decided.93 Perhaps a perfectly just
society would make special efforts to accommodate its religious mino-
rities’ special needs, but international legitimacy does not demand
perfect justice. It only requires the decency of reasons, offered in good
faith, connected to the promotion of the common good.
Cambridge Books
Military service continued

While courts have repeatedly held that “it is not persecution for a
country to require military service of its citizens”94 – even when that
law burdens religious minorities like the Canas brothers – they have at
the same time carved out two exceptions. The first exception involves
cases where punishment is imposed on a particular objector on account
of an illegitimate reason. For example, if a state selectively subjected
Jehovah’s Witnesses to especially severe punishment for failure to serve,
the punishment would be regarded as persecutory.
The second exception encompasses cases where “the alien would
necessarily be required to engage in inhuman conduct as a result of
military service required by the government.”95 Thus, one court has
granted asylum to a Salvadoran who was “ordered by a military officer,
under threat of death, to participate in the paid killing of two men.”96
This second exception is rooted in the anti-brutality principle: interna-
tional humanitarian law places limits on what states may do, and order
their soldiers to do, during battle.*
* It is important to distinguish between violations of jus in bello – the law governing

conduct in war – and violations of jus ad bellum – the law governing the conditions
under which a state may resort to war. While courts have been open to an asylum
seeker’s claim that he would be forced to engage in violations of jus in bello, they have
generally rejected claims that an asylum seeker would be forced to participate in an
unjust war. Just as states have the authority to draft their citizens into military service,
they also have the authority to order those citizens to fight, even if the war is unjust.
There is nothing “inhuman” about fighting an unjust war in accordance with the rules
of international humanitarian law. But states cannot order their soldiers to carry out
atrocities – even if there are just reasons for going to war. Applying that distinction, a
Canadian court rejected an asylum claim by an American soldier who objected to
being forced to serve in Iraq. Hinzman v. Minister of Citizenship and Immigration
[2006] F.C. 420, at 188. (On appeal, the court declined to address that issue, instead
finding that the applicant had not exhausted the protection available to him under the
American exemption for conscientious objectors. Hinzman v. Canada, [2007]
F.C.A. 171.)
The soldier’s liability as a war criminal tracks the distinction between jus ad
bellum and jus in bello: one can become a war criminal by virtue of illegal acts
committed as a foot soldier, that is, for violations of jus in bello; one cannot become
a war criminal merely by having participated in an illegal war. For a general
discussion, see Cecilia M. Bailliet, “Assessing Jus ad Bellum and Jus in Bello Within
the Refugee Status Determination Process: Contemplations on Conscientious
Objectors Seeking Asylum,” Georgetown Immigration Law Journal, 20 (2006),
pp. 337–84.
Cambridge Books
Courts emphasize, however, that to qualify for this second exception,

the inhuman conduct must be ordered as an official policy. “[M]iscon-
duct by renegade military units is almost inevitable during times of
war,”97 and does not constitute official action. That caveat also follows
from asylum’s expressive dimension. Condemnation is misplaced if
atrocities are carried out by renegades against official orders. Of course,
when “renegade” units are common, and the military appears to con-
done their operation, then the atrocities they carry out do reflect official
state policy.
What constitutes the kind of “inhuman conduct” in the course of
military service that can give rise to eligibility for asylum? The UNHCR
Handbook, a guide to the interpretation of the Refugee Convention,
suggests that the conduct must be “condemned by the international com-
munity as contrary to basic rules of human conduct.”98 One court has read
this to mean that “recognized international governmental bodies” (such as
the United Nations or the Organization of American States) must have
actually condemned the conduct.99 Under this standard, a court could not
rely on news clippings or reports by groups like Human Rights Watch or
Amnesty International to establish that conduct was “inhuman.”
That is misguided. An asylum seeker’s claim should not depend on
whether the United Nations or some other intergovernmental organization
has decided to issue a resolution labeling his state’s military policies to be
“inhuman.” International organizations may fail to take such action for
many reasons entirely unrelated to the barbarity of the state’s conduct. To
defer to the judgment of such organizations is an abdication of the judicial
role. “[T]he basic rules of humanitarian conduct are well documented
and readily available to guide [courts] in discerning what types of actions
are considered unacceptable by the world community,”100 and courts are
expert at assessing the strength of documentary evidence submitted to them
in order to determine whether a legal standard has been met.
One court has expressed concern that “[t]his responsibility would
require [courts] to make immigration decisions based on our own
implicit approval or disapproval of U.S. foreign policy and the acts of
other nations. Courts could be put in the position of ruling, as a matter
of law, that a government whose actions have not been condemned by
international governmental bodies engages in persecution against its
citizens.”101 But, I have argued throughout, that is precisely what
asylum law asks adjudicators to do. And, as I suggested in Chapter 2,
concern about judicial meddling in politics is to some degree mitigated
Cambridge Books
by the fact that administrative law judges – who are part of the executive
branch – determine asylum eligibility in the first instance. Although
their determinations are reviewed by the courts, review is deferential
to both the agency’s factfinding and its policy judgments.
Civil war
Alexander Aleinikoff has referred to the case of “persons caught in the
crossfire of a civil war” as “the hypothetical case that is always present in
discussions of the definition of refugee.”102 Civil wars often affect entire
populations, and fear of opening the floodgates has long led US courts to bar
asylum claims for non-combatants on the theory that war is different:
“activities directly related to civil war are not persecution,”103 even if they
involve atrocities that would constitute persecution during peacetime. An
exception has been made for asylum seekers fleeing ethnic or religious
conflicts. Victims of ethnic violence are harmed on account of a Convention
reason, and so fall squarely within the Convention definition.104 Canadian
courts have been more open to applicants fleeing civil wars, extending
asylum so long as applicants could establish that their “fear [was] not that
felt indiscriminately by all citizens as a consequence of the civil war,”105 but
instead was especially acute due to a Convention reason.*
Some have attacked current law as being too narrow, contending that
asylum law should recognize that “aerial bombardments of villages,
‘scorched-earth’ tactics, and pitched battles in urban neighbourhoods
constitute persecution of innocent victims” regardless of the attackers’
motivations and regardless of whether the same risk is generally shared
by others.106 They suggest that the nexus clause should be expanded
to include all individuals fleeing civil strife, and propose that in the
* Many states offer protection on a basis other than asylum to civil war victims who
do not qualify under the Convention definition. In the EU, applicants fleeing a civil
war can qualify for subsidiary protection even when a civil war is not ethnically
based, so long as they can show a “serious and individual threat to life or person
by reason of indiscriminate violence.” EU Qualifications Directive, Council
Directive 2004/83/EC of April 29, 2004, art. 15. In Canada, such applicants are
regarded as “person[s] in need of protection” who are eligible for asylum if they
can show that they would face a risk to their life “in every part of [their] country”
and that the risk “is not faced generally by other individuals in or from that
country,” regardless whether that risk is due to a Convention reason. Immigration
and Refugee Protection Act (2001, c. 27) section 97(1)(b).
Cambridge Books
meantime, flight from civil strife should give rise to a presumption that
the nexus clause has been met.107
How does the political approach to asylum respond to the situation
of civil war? It depends. A distinction must be drawn between, on the
one hand, civil wars that are primarily battles for power between a
government and rebel group, each of which claims to represent the
entire community; and on the other hand, civil wars fought along
ethnic, racial, national, or religious lines.
In the first type of civil war, civilians are aptly described as cross-fire
victims. When villages become the settings for battles, villagers are deprived
of physical security, but their insecurity is a side effect, not the intention, of
the combatants’ operations. When states use force to defend themselves
against insurrection, the tragic and unavoidable consequence is that some
innocent civilians will lose their lives. Frequently, though, civilians can
remove themselves from the battlefield and reside temporarily in another
part of the country. Once even a fragile peace takes hold, or the lines of
battle shift, they will be able to return to their homes to rebuild their lives
free from harm. In that situation, asylum would be inappropriate. If
civilians need international protection for the duration of the conflict, it
can be provided through refugee policy tools other than asylum.
A different scenario is presented if civilians are directly targeted by the
combatants for harm. Clearly, asylum is appropriate in cases of ethnic
conflict and genocide. The deliberate targeting of civilians for harm on
account of their race, religion, or nationality belies any notion that the
responsible combatant acts on behalf of the entire citizenry. Indeed,
asylum is appropriate when combatants target civilians for harm even
absent any Convention reason. For example, combatants may seek to
depopulate villages as a military tactic, without regard to the racial, ethnic,
religious, or political make-up of the villages. In the 1980s, for example,
the Salvadoran government depopulated areas of possible guerrilla resis-
tance through “the use of murder, torture, rape, the burning of crops in
order to create starvation conditions, and a programme of general terror-
ism and harassment.”108 When a government turns on its citizens in such
a manner, its conduct enjoys no presumption of legitimacy; rather, such
conduct constitutes a clear violation of international humanitarian law,
and should be regarded as persecution under the rubric of the anti-
brutality principle. That conclusion could be accommodated within the
present law by presuming that any violation of international humanitar-
ian law is for a Convention reason.109
Cambridge Books
Kleptocracy and negligence

Another hard case involves “kleptocracy” – government by thievery –
characterized by the extortion of citizens by security forces. The case of
Fritz Desir, an asylum applicant from Haiti, presents this problem with
clarity. On several occasions over the period of a few years, Desir was
arrested and beaten by the Ton Ton Macoutes, the Haitian security forces
under the Duvalier regime, for having failed to pay “protection money”
to the Macoutes and for having competed with Macoutes-approved
businesses. The BIA denied his application for asylum, concluding that
the Macoutes “harassed Desir not because of his … political opinion” or
any other ground protected by the nexus clause, but instead “because
they wished to extort money from him for personal reasons.”110
On appeal, the Ninth Circuit Court of Appeals reversed the BIA. Its
reasoning was framed to satisfy the nexus requirement. It noted that
Haiti operated as a “kleptocracy” in which the security forces were
unpaid and depended for their livelihood on corruption and the extor-
tion of their fellow citizens. The reprisals inflicted on those who refused
to pay up were “tactics whereby the Duvalier regime systematically
exercised its authority by way of terror and intimidation.” In light of
this political context, it reasoned, Desir’s “[r]efusal to accede to extor-
tion resulted in his classification and treatment as a [political] subver-
sive.” He was “perceived as disloyal and subversive and the machinery
of the state, enforced by the Macoutes, was violently engaged against
him.” Thus, the court concluded, he was persecuted on account of
political opinion.111
The court’s decision is defensible from the standpoint of legitimacy.
In the interest of promoting the rule of law, states are increasingly
imposing sanctions on members of foreign governments who are
responsible for kleptocracy and corruption. Such sanctions range
from the withholding of aid to the freezing of assets. When a citizen is
subjected to serious harm for refusing to capitulate to extralegal extor-
tion, international protection and international condemnation are both
warranted. Asylum is an appropriate response.
But what about the indirect victims of kleptocracy, the people
reduced to a marginal existence due to economic policies that are
designed to enrich and entrench those in power? Can they too claim
asylum? This presents a much closer case. At least one observer has
Cambridge Books
argued that “where the political system serves only to perpetuate eco-
nomic disparity and widespread impoverishment,” courts should recog-
nize the existence of “structural economic persecution”112 and award
asylum to those fleeing impoverished conditions. Much can be said for
that position. Kleptocratic economic policies are deserving of condem-
nation, and if the impoverishment and suffering they cause (albeit
indirectly) rise to the level of “serious harm,” asylum would seem to
be a fitting response.
On the other hand, however, describing indirect harm as “persecu-
tion” seems to stretch the concept beyond its language. The term con-
notes targeting or selecting particular individuals, or a particular class
of individuals, for harm. Dictionary definitions of the term “persecu-
tion” employ synonyms like “pursue,” “torment,” “harass,” “punish,”
and “exterminate.” The word “persecution” does not describe accu-
rately capture the situation of people impoverished by kleptocratic
states. Their suffering may be traceable to illegitimate policies, but it is
not the intention of those policies.
In the end, a concern for practical consequences points against grant-
ing asylum to victims of “structural economic” harm. Much of the
world could arguably be described as falling into that category.
Furthermore, the line between confiscatory policies worthy of condem-
nation and inept policies deserving of international assistance can often
be hard to draw. Developing nations often seek to promote the rule of
law, but find that embedded interests make success difficult to achieve.
For similar reasons, asylum is an inappropriate response to the threat
of harm caused by official negligence – for example, a state’s failure to
take adequate steps to protect citizens against environmental cata-
strophe. The condemnation expressed by asylum is more appropriate
when harm is intended than when it is the unintended consequence of
negligence. Also, it may be quite difficult for adjudicators to determine
whether a state has been negligent. The concept of negligence is depen-
dent on some notion of a reasonable standard of care; but what is
reasonable will vary widely across societies, depending on the available
economic resources, competing priorities which lay claim to those
scarce resources, and practical challenges – such as a lack of human
capital, poor infrastructure, and barriers to communication, command,
and control – that make it difficult to execute even a well-conceived
policy. Moreover, technical experts may often disagree about whether,
for example, a particular economic policy is wise or so foolish as to be
Cambridge Books
negligent. By contrast, in the garden variety asylum case involving a

citizen targeted for harm, these problems do not arise.
“Persecution” also does not generally describe the situation of famine
victims, even though, as Amartya Sen has famously argued, famine is
the consequence of a political and economic failure, that is, the lack of
an entitlement to food, not a lack of available food.113 In the usual case
of famine, no one intended for the victims to starve; their situation is the
result of inept governance, not malice. The appropriate posture toward
the victims’ state is assistance, not condemnation.
Of course, there are exceptions. An entitlement to food can some-
times be withheld for an illegitimate reason. For example, in Zimbabwe
in 2005, limited grain supplies were funneled to areas expected to be
“swing” areas in upcoming elections. Those regions which were either
solidly behind the ZANU-PF ruling party or the opposition Movement
for Democratic Change thus bore the brunt of the famine caused by the
mismanaged land reform of the 1990s. Furthermore, even in swing
areas, the state-controlled Grain Marketing Board would not sell corn
to rural families without a ZANU-PF membership card. And President
Robert Mugabe refused to allow the World Food Program to deliver
food to these areas, because to allow international assistance would
belie his claim that the land reform had resulted in a bumper harvest.114
In cases like that of Zimbabwe, asylum is an appropriate response
to famine.
Conclusion
I have argued that “persecution” should be defined as serious harm
inflicted or condoned by official agents for illegitimate reasons. The
“serious harm” requirement recognizes that asylum is not meant to
remedy every wrong or injustice committed by a state against a citizen;
harm must reach a certain threshold of seriousness before it warrants
international condemnation and external interference. The reference to
“illegitimate reasons” refers to reasons that belie the principle of com-
munal self-determination that justifies the ordinary presumption of
legitimacy given to state policies. Asylum policy involves a series of
judgments about what reasons fall within that category and thus
warrant international condemnation and external interference. The
nexus clause provides some guidance, but it is not exclusive. Indeed,
certain types of harm – such as torture and other violations of
Cambridge Books
peremptory international human rights norms – are always beyond the

pale, no matter why they have been inflicted.
Asylum law is therefore an ongoing normative enterprise – one in
which adjudicators draw and articulate the bounds of legitimate state
conduct in light of concrete cases and scenarios. Some might object
that adjudicators will rely too much on personal, idiosyncratic views
about what states may and may not do to their citizens. But that
concern is overstated. A number of mechanisms exist to channel
adjudicators’ discretion. In addition to the Convention definition
itself, hard calls can be made collectively through legislation or
regulations that govern an entire class of cases; other times, an adjudi-
cator’s decision is controlled by the accretion of legal precedent
reflecting a consensus borne of experience. But sometimes, an adjudi-
cator will face novel issues and will need to consult his or her own
judgment.
Some have suggested that international human rights law should
provide the measure of legitimacy, and that the term “persecution”
should be interpreted in light of the rights delineated in the international
human rights covenants. Such an approach has been endorsed by some
of the Commonwealth countries. But, for a number of reasons, that
approach masks difficult normative determinations. First, not all viola-
tions of human rights constitute persecution; second, human rights
themselves must be interpreted and applied to concrete factual situa-
tions in light of underlying, and too often unarticulated, notions of
legitimacy; third, human rights can conflict, requiring an adjudicator
to resort to a higher order principle, namely, legitimacy. In sum, while
the roster of human rights can offer some guidance, in the end asylum
adjudicators and policymakers will need to engage directly with under-
lying first principles of legitimacy.
Some may object that the approach to asylum advanced here leaves
too many needy people unprotected. But that humanitarian objection
misunderstands asylum’s role in the refugee policy toolkit. Asylum is
not the one-size-fits-all solution for all of the world’s refugees. It has a
more targeted focus, connected to its expressive character. At the same
time, however, states should also take seriously their humanitarian
obligation to assist refugees who are excluded by asylum, using refugee
policy tools like relief aid, development assistance, and humanitarian
protection abroad (typically on a temporary basis). A targeted asylum
policy does not justify a stingy refugee policy.
Cambridge Books
Notes
1. UNHCR, Handbook on Procedures and Criteria for Determining Refugee
Status Under the 1951 Convention and the 1967 Protocol Relating to the
Status of Refugees, UN Document HCR/PRO/4 (1979), revised 1992,
paragraph 51.
2. Guy Goodwin-Gill, The Refugee in International Law, 2nd edn. (New
York: Oxford University Press, 1996), p. 67.
3. The US courts of appeals have employed a number of definitions of persecu-
tion which, if carefully parsed, have very different implications for certain
kinds of cases. One widely cited definition treats persecution as “the inflic-
tion of suffering or harm, under government sanction, upon persons who
differ in a way regarded as offensive … in a manner condemned by civilized
governments.” Kovac v. INS, 407 F.2d 102, 107 (9th Cir. 1969). This
definition would exclude harm committed by non-state actors without
government sanction. Another widely cited definition states that persecution
is “harm or suffering … inflicted upon an individual in order to punish him
for possessing a belief or characteristic a persecutor sought to overcome.”
Matter of Acosta, 19 I. & N. Dec. 211, 223 (BIA 1985). This definition
suggests that, to be classified as persecution, harm must be inflicted with a
punitive intent. That would exclude, for example, the victim of female
genital mutilation or a lesbian who was subjected to forced institutionaliza-
tion to “cure” her of her lesbianism. Cf. Pitcherskaia v. INS, 118 F.3d 641
(9th Cir. 1997). Another widely cited definition states that persecution is
“oppression which is inflicted on groups or individuals because of a differ-
ence that the persecutor will not tolerate.” Hernandez-Ortiz v. INS, 777
F.2d 509, 516 (9th Cir. 1985). That definition suggests that persecution
could result from the state’s failure to exempt, say, a religious group from a
generally applicable law that severely burdens the religious group. The
failure to accommodate could be regarded as a refusal to “tolerate” a
“difference.” The BIA has more recently defined persecution as “the inflic-
tion of harm or suffering by a government, or persons a government is
unwilling or unable to control, to overcome a characteristic of the victim.”
Matter of Kasinga, 21 I. & N. Dec. 357, 365 (BIA 1996). But the older
definitions just reviewed nonetheless continue to be cited by courts.
4. See note 3, above.
5. R. v. Secretary for the Home Department, ex parte Shah [1999] 2 A.C.
653; Minister for Immigration and Multicultural Affairs v. Khawar [2002]
210 C.L.R. 1, at 120; New Zealand Refugee Appeal No. 74665/03 (July 7,
2004), paragraph 53.
6. Michael Walzer, Just and Unjust Wars (New York: Basic Books, 1977),
p. 54.
Cambridge Books
7. Samuel Pufendorf, De Jure Naturae et Gentium Libri Octo (1672), trans.

C. H. Oldfather and W. A. Oldfather (Oxford: Clarendon Press, 1934),
VIII.6.14.
8. Walzer, Just and Unjust Wars, p. 89.
9. Ibid., p. 90.
10. Ibid., p. 101.
11. Ghaly v. INS, 58 F.3d 1425, 1431 (9th Cir. 1995).
12. Fatin v. INS, 12 F.3d 1233, 1240 (3rd Cir. 1993).
13. Mikhailevitch v. INS, 146 F.3d 384, 390 (6th Cir. 1998).
14. Aguilar-Solis v. INS, 168 F.3d 565, 570 (1st Cir. 1999).
15. Begzatowski v. INS, 278 F.3d 665, 669 (7th Cir. 2002), quoting Mitev v.
INS, 67 F.3d 1325, 1330 (7th Cir. 1995).
16. Li v. Ashcroft, 312 F.3d 1094 (9th Cir. 2002), vacated and rev’d, Li v.
Ashcroft, 356 F.3d 1153 (9th Cir. 2004) (en banc).
17. Ibid., p. 1098.
18. Prasad v. INS, 47 F.3d 336 (9th Cir. 1995).
19. Li, 312 F.3d at 1101.
20. Ibid., p. 1105.
21. Aguilar-Solis, 168 F.3d at 570.
22. Diallo v. Ashcroft, 381 F.3d 687, 697 (7th Cir. 2004); Osaghae v. INS,
942 F.2d 1160, 1163 (7th Cir. 1991).
23. See, for example, Matter of Izatula, 20 I. & N. Dec. 149 (BIA 1990);
Dwomoh v. Sava, 696 F. Supp. 970 (S.D.N.Y. 1988); Carolyn Patty
Blum, “License to Kill: Asylum Law and the Principle of Legitimate
Government Authority to ‘Investigate its Enemies,’” Willamette Law
Review, 28 (1992), p. 719; Carolyn Patty Blum, “Political Assumptions in
Asylum Decision-Making: The Example of Refugees from Armed Conflict”
in Howard Adelman (ed.), Refugee Policy: Canada and the United States
(Toronto: York Lanes Press, 1991), pp. 282–91; Note, “Political Legitimacy
in the Law of Political Asylum,” Harvard Law Review, 99 (1985), p. 450.
24. Matter of Acosta, 19 I. & N. 211, 233 (BIA 1985).
25. Immanuel Kant, “On the Common Saying, ‘This May Be True in Theory,
But It Does Not Apply in Practice’” in Immanuel Kant, Political Writings,
trans. and ed. Hans Reiss (Cambridge University Press, 1991), pp. 84–5
(emphasis in original).
26. Daniel J. Steinbock, “Interpreting the Refugee Definition,” UCLA Law
Review, 45 (1998), p. 794.
27. American Law Institute, Restatement of the Law, Third: The Foreign
Relations Law of the United States (St. Paul, MN: American Law
Institute Publishers, 1987 & Supp. 2007), section 702(a)–(f) and cmt.n.
28. EU Qualifications Directive, Council Directive 2004/83/EC of April 29,
2004 on minimum standards for the qualification and status of third
Cambridge Books
country nationals or stateless persons as refugees or as persons who

otherwise need international protection and the content of the protection
granted, art. 15.
29. Matter of R –, 20 I. & N. Dec. 621, 622 (BIA 1992).
30. Ibid., pp. 623–5.
31. See, for example, T. Alexander Aleinikoff, “The Meaning of ‘Persecution’
in U.S. Asylum Law” in Adelman (ed.), Refugee Policy: Canada and the
United States (Toronto: York Lanes Press, 1991), pp. 309–13; Michelle
Foster, “Causation in Context: Interpreting the Nexus Clause in the
Refugee Convention,” Michigan Journal of International Law, 23
(2002), p. 338.
32. 8 C.F.R. section 208.16.
33. INS v. Cardoza-Fonseca, 480 U.S. 421, 431 (1987).
34. EU Qualifications Directive, art. 24.
35. In re S – P – , 21 I. & N. Dec. 486, 494 n. 3 (BIA 1996).
36. Aleinikoff, “The Meaning of ‘Persecution,’” p. 309.
37. For discussion, see Deborah E. Anker, “Refugee Law, Gender, and the
Human Rights Paradigm,” Harvard Human Rights Journal, 15 (2002),
pp. 133–54; Karen Musalo and Stephen Knight, “Steps Forward and
Steps Back,” International Journal of Refugee Law, 13 (2001), pp. 51–
70; and Erik D. Ramanthan, “Queer Cases: A Comparative Analysis of
Global Sexual Orientation-Based Asylum Jurisprudence,” Georgetown
Immigration Law Journal, 11 (1996), pp. 1–44.
38. 8 U.S.C. section 1101(a)(42).
39. James Hathaway has been the most influential advocate of a tight linkage
between refugee law and international human rights law. See James C.
Hathaway, Law of Refugee Status (Toronto: Butterworths, 1991); James
C. Hathaway, “Reconceiving Refugee Law as Human Rights
Protection,” Journal of Refugee Studies, 4 (1991), p. 113. Courts in
Canada and New Zealand have adopted his analysis in order to deter-
mine eligibility for asylum in their respective countries. See Canada
(Attorney General) v. Ward [1993] S.C.R. 689; New Zealand Refugee
Status Appeals Authority, Refugee Appeal No. 71427/99 [2000]
N.Z.A.R. 545. He is quoted approvingly in British and Australian asylum
decisions as well. See, for example, Horvath v. Secretary of State for the
Home Department [2001] 1 A.C. 489 (UK); and Minister for
Immigration and Multicultural Affairs v. Khawar [2002] H.C.A. 14
(Australia). The linkage between international human rights law and
asylum law is less developed in the United States.
40. Hathaway, Law of Refugee Status, p. 106.
41. Hathaway, “Reconceiving Refugee Law,” p. 123.
42. Hathaway, Law of Refugee Status, pp. 104–5.
Cambridge Books
43. Ibid., pp. 109–11.

44. Hathaway states, for example, that “[u]nder current interpretations,
refugee status requires a risk to basic human rights … in addition to
some differential impact based on civil or political status [i.e. a
Convention reason] … The proposal here is that refugee status become
the entitlement of all persons whose basic human rights are at risk.”
Hathaway, “Reconceiving Refugee Law,” p. 121 (emphasis in original).
45. International Covenant on Civil and Political Rights, December 16,
1966, 999 U.N.T.S. 171 (entered into force March 23, 1976), art. 25
[ICCPR].
46. International Covenant on Economic, Social and Cultural Rights,
December 16, 1966, 933 U.N.T.S. 3, art. 9.
47. Refugee Appeal No. 74665/03 (July 7, 2004) (New Zealand), paragraph
53, available at www.nzrefugeeappeals.govt.nz. The formula quoted
above was originally expressed in the UK decision ex parte Shah [1999]
2 A.C. at 653; it was also approved of by the Australian High Court in
Khawar [2002] HCA 14, at 120.
48. Universal Declaration of Human Rights, G. A. Res. 217A, U.N. GAOR,
3rd Sess., Supp. No. 3, at 71, U.N. Doc. A/810 (1948), art. 17.
49. See, for example, Begzatowski, 278 F.3d at 669.
50. German Extradition Law of December 23, 1929, art. 3(2), quoted in
Dwomoh, 696 F. Supp. at 976 n. 8.
51. ICCPR, art. 19(2)–(3).
52. Fisher v. INS, 79 F.3d 955, 959–60 (9th Cir. 1996) (en banc).
53. ICCPR, art. 19(2).
54. Ibid., art. 19(3)(b).
55. Thanks to Inken Wiese for this example.
56. This argument assumes that the punishment for violating the chador law
is appropriate for the violation of a morals offense. Evidence in some
chador cases suggests that activists who violate the law are sentenced to
up to one year in prison and subjected to public corporal punishment in
the form of lashings. The harshness of that sentence – especially if it is
only imposed on activists – raises an inference that although the formal
charge is for violating the chador law, the punishment is inflicted on
account of political opinion.
57. Steinbock, “Interpreting the Refugee Definition,” p. 755.
58. ICCPR, art. 26.
59. Fisher, 79 F.3d at 961–2 (alterations in original).
60. The chador case also offers another example of the way in which a human
rights approach can be overinclusive. Even if one were to conclude that
the chador law was discriminatory and therefore in violation of the
ICCPR, one might still conclude that the discrimination was not
Cambridge Books
sufficiently serious to rise to the level of persecution. The chador law may
be “inconvenient, irritating, mildly objectionable, or highly offensive,”
but it does not mandate the infliction of physical pain or harm (unlike,
say, the hypothetical foot-binding law or like female genital mutilation),
and courts generally conclude that discrimination or minor harassment is
not sufficiently serious to constitute persecution – even though such
discrimination or harassment can cause mental trauma, as in Hassib-
Tehrani’s case. Fisher, 79 F.3d at 962. See also Yadegar-Sargis v. INS,
297 F.3d 596 (7th Cir. 2002); Fatin, 12 F.3d 1233; Safaie v. INS, 25 F.3d
636 (8th Cir. 1994).
61. “Liberty and equality, freedom and security, private property and dis-
tributive justice … conflict, and, because they do, the rights that define
them as entitlements are also in conflict …” Michael Ignatieff, Human
Rights as Politics and Idolatry (Princeton, NJ: Princeton University Press,
2001), pp. 20–1.
62. Matter of Chang, 20 I. & N. Dec. 38 (1989). Currently, the Immigration
and Nationality Act provides that “a person who has been forced to abort
a pregnancy or to undergo involuntary sterilization, or who has been
persecuted for failure or refusal to undergo such a procedure or for other
resistance to a coercive population control program, shall be deemed to
have been persecuted on account of political opinion, and a person who
has a well founded fear that he or she will be forced to undergo such a
procedure or subject to persecution for such failure, refusal, or resistance
shall be deemed to have a well founded fear of persecution on account of
political opinion.” 8 U.S.C. section 1101(a)(42)(B).
63. While “one couple, one child” was the general rule, exceptions were
made in limited circumstances, for instance if the first child was disabled
and so unable to work. Matter of Chang, 20 I. & N. Dec. at 40.
64. Ibid., p. 39.
65. Ibid., p. 47.
66. See, for example, Jacqueline Bhabha, “Embodied Rights: Gender
Persecution, State Sovereignty, and Refugees,” Public Culture, 9
(1996), p. 11.
67. Cheung v. Canada (Minister of Employment and Immigration) [1993]
102 D.L.R. (4th) 214, 221–22 (Fed. Ct. App.).
68. Matter of Chang, 20 I. & N. Dec. at 43–4.
69. One might offer the following rejoinder: “Socio-economic human rights
are abrogated only where a state either neglects their realization in the
face of adequate resources, or implements them in a discriminatory
way; … the existence of generalized hardship is [not] a sufficient … factor
in defining the existence of socio-economic persecution.” Hathaway,
Law of Refugee Status, p. 119. Thus, China faces no conflict of rights.
Cambridge Books
Although it might be unable to provide for a larger population, this

inability arises from neither neglect nor discrimination. That response is
unconvincing. Subsistence is certainly a critical component of human
dignity and an interest of such fundamental importance that China
should seek to remedy the situation with the same urgency as it should
seek to remedy violations of civil and political rights.
70. Matter of Maldonado-Cruz, 19 I. & N. Dec. 509, 518 (BIA 1988). The
BIA’s decision was overturned by the Ninth Circuit on the ground that
Maldonado-Cruz’s neutrality amounted to a political opinion.
Maldonado-Cruz v. U.S. I.N.S., 883 F.2d 788 (9th Cir. 1989).
However, other circuits declined to follow the Ninth Circuit in holding
that political neutrality is a political opinion. See, for example, Perlera-
Escobar v. Executive Office for Immigration, 894 F.2d 1292, 1297 n. 4
(11th Cir. 1990).
71. Blum, “Political Assumptions,” offers a similar argument.
72. The BIA made a second mistake in the quote above as well: it confused the
legitimacy of an end with the legitimacy of the means. Even if the
Salvadoran government did have a right to defend itself against rebellion,
it would not follow that the government’s legitimate right to investigate
and detain its enemies extended to the treatment feared by Maldonado-
Cruz. A government’s general right to repress uprisings does not justify
any action taken toward that end. The anti-brutality principle places a
substantive limit on the steps a state may take to defend itself.
Another example of this second mistake was provided by another
Board decision, involving Indian counter-terror operations. In In Re
Jagraj Singh, the BIA denied eligibility for asylum to an Indian Sikh
who was harassed and tortured by police after providing food to Sikh
militants who had threatened his life if he refused to help them. The police
“beat [Singh] repeatedly with a rifle, and asked him about the militants.
When Singh did not reveal the militants’ names, the police shot him in the
leg.” On another occasion, the police beat Singh so severely that he could
not walk for two weeks, and he was only released when his mother gave a
$25,000 bribe. Unpublished opinion cited and overturned in Singh v.
Ilchert, 801 F. Supp. 313, 316 (N.D. Cal. 1992). The BIA ruled that Singh
was beaten as part of a criminal investigation falling under Indian anti-
terrorism laws, and that the police were interested in gaining information
about Sikh militants, not in punishing Singh for his political views. The
police, the Board said, have a “legitimate right to investigate to determine
whether [Singh] could provide information about these people.” Ibid.,
p. 317. For other examples of the BIA’s endorsement of a “legitimate
right” to self-defense, see Blum, “License to Kill,” pp. 729 et seq. Asylum
adjudication should consider not only whether a government may
Cambridge Books
legitimately defend itself, but also whether the means employed in pursuit
of that end are legitimate. See Blum, “Political Assumptions” and Note,
“Political Legitimacy” for a similar point.
73. Dwomoh, 696 F. Supp. at 979 and n. 11. Even assuming arguendo that
Ghana did have a legitimate right to punish Dwomoh, the court contin-
ued, it needed to do so in accordance with due process. Ibid., p. 978. See
also Izatula, 20 I. & N. Dec. 149.
74. Chanco v. INS, 82 F.3d 298 (9th Cir. 1996).
75. Sir James Fitzjames Stephens, History of the Criminal Law of England,
vol. 2 (London: Macmillan, 1883), p. 70.
76. See, for example, INS v. Aguirre-Aguirre, 526 U.S. 415, 421–2 (1999); 8
U.S.C. section 1231(b)(3)(B); UN Convention Relating to the Status of
Refugees, July 28, 1951, 189 U.N.T.S. 137, art. 1(F)(b).
77. Matter of R –, 20 I. & N. Dec. at 635 (Heilman, Board member,
concurring).
78. Ibid., p. 629 (majority opinion).
79. Hugo Grotius, De Jure Belli ac Pacis Libri Tres (1625), trans. Francis W.
Kelsey (Oxford: Clarendon Press, 1925), II.21.5.
80. Emer de Vattel, The Law of Nations or the Principles of Natural Law
(1758), trans. Charles G. Fenwick (Geneva: Slatkine Reprints – Henry
Dunant Institute, 1983), I.19.232.
81. Thomas Jefferson, “Letter to Washington, Nov. 7, 1791,” in Thomas
Jefferson Randolph (ed.), Memoir, Correspondence, and Miscellanies
from the Papers of Thomas Jefferson, vol. 3, 2nd edn. (Boston, MA:
Gray and Bowen, 1830), p. 131.
82. 8 C.F.R. section 1208.16; 8 U.S.C. section 1231(b)(3)(B).
83. 8 U.S.C. section 1158(b)(2)(A).
84. EU Qualifications Directive, art. 12; Immigration and Refugee Protection
Act (2001, c. 27), section 98 (Canada); UN Convention, art. 1(F).
85. Castaneda-Castillo v. Gonzales, 488 F.3d 17, 20 (1st Cir. 2007) (en banc).
86. Hernandez v. Ashcroft, 345 F.3d 824 (9th Cir. 2003).
87. 8 U.S.C. section 1182(a)(3)(B), (F).
88. Human Rights First, “Abandoning the Persecuted: Victims of Terrorism
and Oppression Barred From Asylum” (2006), www.humanrightsfirst.
info/pdf/06925-asy-abandon-persecuted.pdf (last visited February 27,
2008).
89. Department of Homeland Security, “Exercise of Authority Under Sec.
212(d)(3)(B)(i) of the Immigration and Nationality Act,” 72 Fed. Reg.
9958 (March 6, 2007).
90. Canas-Segovia v. INS, 902 F.2d 717, 723 (9th Cir. 1990). After the Ninth
Circuit reversed, holding that the Canas brothers had been persecuted on
account of religion, 902 F.2d at 725–6, the US Supreme Court vacated
Cambridge Books
the decision, 502 U.S. 1086 (1992), based on the rule that, for harm to
constitute persecution on account of religion, the persecutor must have
been motivated to inflict the harm on account of the victim’s religion.
There had been no evidence before the court that the Salvadoran govern-
ment had refused to allow an exemption for Jehovah’s Witnesses because
it sought to harm them. Rather, the government was likely simply indif-
ferent to the fact that some religious minorities might be severely bur-
dened by compulsory conscription. On remand in the Canas brothers’
case, the Ninth Circuit found persecution on grounds of imputed political
opinion. Canas-Segovia v. INS, 970 F.2d 599, 602 (9th Cir. 1992).
91. Foroglou v. INS, 170 F.3d 68, 71 (1st Cir. 1999).
92. Canas-Segovia, 902 F.2d at 723.
93. Zehatye v. Gonzales, 453 F.3d 1182 (9th Cir. 2006); Foroglou, 170
F.3d 68.
94. Matter of A – G –, 19 I. & N. Dec. 502, 506 (BIA 1987).
95. Ibid.
96. Barraza Rivera v. INS, 913 F.2d 1443, 1450 (9th Cir. 1990).
97. M.A. v. INS, 899 F.2d 304, 312 (4th Cir. 1990).
98. UNHCR, Handbook, section 171.
99. M.A., 899 F.2d at 313.
100. Ibid., p. 323 (Winter, J., dissenting).
101. Ibid., p. 313 (majority opinion).
103. Matter of Rodriguez-Majano, 19 I. & N. Dec. 811 (BIA 1988). See also
UNHCR, Handbook, paragraph 164, stating that “[p]ersons compelled
to leave their country of origin as a result of international or national
armed conflicts are not normally considered refugees under the 1951
Convention or 1967 Protocol.”
104. See, for example, In re H –, 21 I. & N. Dec. 337 (BIA 1996); Minister for
Immigration & Multicultural Affairs v. Abdi [1999] 162 A.L.R. 105
(Fed. Ct. Aust.); New Zealand Refugee App. No. 71462/99 (September
27, 1999).
105. Salibian v. Minister of Employment & Immigration [1990] 3 F.C. 250,
258 (Canada). See also Adan v. Secretary of State for the Home
Department [1999] 1 A.C. 293 (UK).
107. Ibid., p. 309.
108. Mark R. von Sternberg, “The Plight of the Non-Combatant in Civil
War and the New Criteria for Refugee Status,” International Journal of
Refugee Law, 9 (1997), p. 188 (internal quotation marks omitted).
109. See ibid., p. 195, and Hugo Storey and Rebecca Wallace, “War and Peace
in Refugee Law Jurisprudence,” American Journal of International Law,
Cambridge Books
95 (2001), p. 349. See also: In re S – P – , 21 I. & N. Dec. 486, 494 n. 3

(BIA 1996), stating that “the severity of the violations of the Geneva
Convention may support an inference that the abuse is grounded in one of
the protected grounds under the asylum law.”
110. Desir v. Ilchert, 840 F.2d 723, 725 (9th Cir. 1988).
111. Ibid., pp. 727–9.
112. Note, “Political Legitimacy,” p. 463.
113. Amartya Sen, Poverty and Famines: An Essay on Entitlement and
Deprivation (Oxford: Clarendon Press, 1981).
114. Michael Wines, “Zimbabwe Extends Crackdown on Dissent as Election
Looms,” New York Times, December 24, 2004.
Cambridge Books

07.0 PP 103 145 What Is Persecution

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

07.0 PP 103 145 What Is Persecution

Uploaded by

Copyright:

Available Formats

3 What is “Persecution”?

in the United States, “government sanction” was initially regarded as a

Persecution and international legitimacy

the principle of communal self-determination, the idea that a political

The effect is to permit states to adopt a wide range of domestic institu-

boundaries is so terrible that it makes talk of community or self-

In such cases, external interference advances rather than impinges on

The elements of “persecution”

Li’s forced examination was significantly less serious than Prasad’s

political opinion” – illustrates illegitimate reasons for inflicting serious

The idea that political opinion is an illegitimate basis for inflicting

group membership, or political opinion. The list of grounds given by the

In other words, R failed to qualify for asylum because the police

coercive sterilization program, and a similar statutory presumption was

not persecutory because, as a law of general application, it did not target

The role for human rights

protection elsewhere.”41 Thus, Hathaway concludes, “persecution may

Overinclusion and underinclusion

Human rights are not self-defining

socio-economic rights, which are violated only when a state fails to

disfiguring foot-binding as essential to decency. A law requiring women

Human rights in conflict

The BIA rejected Chang’s application. The population control law, it

China has adopted a policy whose stated objective is to discourage births

roster of human rights given by the international covenants is an unreli-

The case of rebels and other political offenders

resistance and overthrow. Indeed, a limited right to revolution lies at the

alien would not face punishment.”74 A coup is defensible only when

deserving of condemnation. But are his own deeds so odious that he

unwittingly ferries a killer to the site of a massacre can hardly be labeled

Laws of general application

the Canas brothers would certainly lose. A general conscription law is

Military service continued

* It is important to distinguish between violations of jus in bello – the law governing

Courts emphasize, however, that to qualify for this second exception,

Kleptocracy and negligence

negligent. By contrast, in the garden variety asylum case involving a

peremptory international human rights norms – are always beyond the

7. Samuel Pufendorf, De Jure Naturae et Gentium Libri Octo (1672), trans.

country nationals or stateless persons as refugees or as persons who

43. Ibid., pp. 109–11.

Although it might be unable to provide for a larger population, this

95 (2001), p. 349. See also: In re S – P – , 21 I. & N. Dec. 486, 494 n. 3

You might also like